J. Dairy Sci. 102:164–176 https://doi.org/10.3168/jds.2018-14964 © 2019, The Authors. Published by FASS Inc. and Elsevier Inc. on behalf of the American Dairy Science Association®. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Influence of -making recipes on the composition and characteristics of -type cheese

Danton Batty, Joy G. Waite-Cusic,* and Lisbeth Meunier-Goddik Department of Food Science and Technology, Oregon State University, Corvallis 97331

ABSTRACT on these quality metrics the shelf-life of these recipes was estimated with the lactic curd having the shortest, Bloomy rind , including Camembert and re- and the stabilized curd having the longest. Examin- lated varieties, can be produced using alternative pro- ing Camembert-type cheese quality metrics for these 5 cesses that vary based on preacidification, cutting, varieties can assist cheesemakers during recipe formula- curd handling, and ripening parameters. Modification tion and selection of cheese-making practices to achieve of these parameters creates distinct cheeses such as lac- optimum product quality. tic curd, stabilized curd, and hybrids of the two. The Key words: -ripened cheese, stabilized cheese, objective of this study was to determine the influence shelf-life of 5 Camembert-type cheese recipes on the composition and characteristics during ripening. Five varieties of Camembert-type cheese were produced: (1) lactic curd, INTRODUCTION (2) sweet curd, (3) washed curd, (4) solubilized curd, Soft, bloomy rind cheeses, predominantly Cam- and (5) stabilized curd. Cheeses were aged at 13°C for embert- and -type cheeses, compose a small, but 10 d, during the mold growth phase, and 7°C from d 11 significant, segment of the total cheese market in the until 50. Key quality metrics including texture develop- United States. Forty-seven percent of the cheesemakers ment, pH (center and surface), and color were moni- within the American Cheese Society produce surface tored throughout shelf-life. Compositional evaluation mold-ripened cheeses (American Cheese Society, 2016; (d 5; fat, protein, moisture, salt, and minerals) grouped Rebecca Orozco, American Cheese Society, Denver, cheeses into 3 categories: (1) lactic curd, (2) sweet and CO, personal communication). In the past decade, washed curd, and (3) solubilized and stabilized curd. specialty cheese production volumes greatly increased. The lactic curd and stabilized curd were consistently From 2005 to 2015, the annual production of specialty the most different varieties for composition and quality cheese in Wisconsin (the leading state of specialty metrics. Moisture content of Camembert-type varieties cheese production) doubled from 164 to 328 million kg ranged from 53.15 to 57.99%, Ca ranged from 0.23 to (Wisconsin Milk Marketing Board, 2017). Grocery sales 0.45%, and P ranged from 0.21 to 0.40%. All varieties for Camembert- and Brie-type cheeses increased by followed the expected pH evolution on the rind and 4.7% from 2004 to 2005, totaling 4.9 million kg annu- in the paste with the pH of the rind reaching 7 by ally (Buragas, 2006). This increase in availability and d 10, and paste pH reaching 7 between 35 and 50 d. sales of specialty cheese is in part due to the increasing The displacement of the paste (distance traveled upon number of artisan cheesemakers. In Oregon alone, the cutting) for the lactic curd was the greatest among the number of artisan specialty cheese processors has expe- 5 varieties, reaching an average of 27 ± 1.9 mm (mean rienced significant growth from 3 in 1999 to 26 in 2013 ± standard error) after 50 d of ripening and 60 min (Bouma et al., 2014). An increase has also occurred of flow time. The stabilized curd on the other hand on a national level, with the American Cheese Society traveled the shortest distance, reaching an average of (2018) reporting a membership increase from 1,418 to 4 ± 0.4 mm at the same time point. Browning, consid- 1,831 over the last decade (2007 to 2017). ered a defect in mold-ripened cheeses, was observed in Soft ripened cheeses were traditionally manufactured all varieties, but was most substantial for lactic curd in using basic soft cheese-making practices (lightness, L*, decreased from 87.19 to 68.58). Based (Shaw, 1981). Key characteristics for soft ripened cheese processes are the fermentation of cheese milk by meso- philic lactic acid bacteria to a low pH (<4.6–5.2) with Received April 21, 2018. Accepted September 12, 2018. an absence of both cooking and pressing steps. This *Corresponding author: joy.waite​-cusic@​oregonstate​.edu category of cheeses can be broken into 2 subcategories

164 COMPOSITION AND CHARACTERISTICS OF CAMEMBERT-TYPE CHEESE 165 based on rind development: washed rind and bloomy The objective of this study was to investigate the rind. Camembert and Brie belong in the bloomy rind influence of 5 commonly used cheese-making recipes on cheese category. the composition, quality characteristics, and shelf-life Camembert and related bloomy rind varieties are of Camembert-type cheese. These 5 varieties capture known for possessing a white/gray rind that is formed many of the procedural differences used to manufacture as candidum, and yeasts such as Geotrichum bloomy rind cheeses in the industry. We hypothesize candidum, Debaryomyces hansenii, and Kluyveromyces that these 5 common recipes will produce Camembert- spp. grow on the surface of the cheese (Gripon, 1997; type cheeses that differ in composition and rate of Leclercq-Perlat, 2011; Galli et al., 2016). Soft ripened ripening as indicated by reduction in firmness and cheeses that undergo transformation by surface stability of the cheese paste. The goal of this study are unique because of the many physiochemical and was to provide recipe information and Camembert-type biochemical changes that occur over the brief period cheese characteristics of each variant to assist artisan of ripening (typically 3–5 wk). These reactions trans- cheesemakers in appropriate recipe selection to achieve form a chalky, crumbly, and firm cheese into one that optimum quality and extend shelf-life as appropriate is soft, viscous, and flowing (Lawrence et al., 1987; for their distribution channels. Sousa, 2003; Tansman et al., 2017). Cheese softening for surface-ripened varieties can be partly explained by MATERIALS AND METHODS mineral migration from the center to the surface of the cheese. This results in the swelling and hydration of Overall Experimental Design the protein matrix (Lucey and Fox, 1993). The mineral migration phenomenon is primarily a result of the pH Five different Camembert-type cheese varieties (lac- increase caused by enzymatic activity from the surface tic curd, sweet curd, washed curd, solubilized curd, molds (Le Graet et al., 1983; Karahadian and Lindsay, and stabilized curd) were manufactured. Differences in 1987; Tansman et al., 2017). manufacturing procedure for the 5 varieties are sum- The optimum time frame for consumption of Cam- marized in Table 1. Each variety was manufactured embert-type cheese is between 3 and 7 wk postproduc- on 2 different days (n = 2). Two cheese wheels from tion (Galli et al., 2016). Therefore, getting the cheese each cheese make were sampled for composition and 3 to market and consumers is a significant challenge for wheels were sampled for physical characteristics at each manufacturers of traditional Camembert-type cheeses time point for analysis. with large distribution networks. Camembert-type cheese can be consumed outside of this optimum qual- Cheese Making ity window; however, defects will be noticeable and often unpleasant. These defects include (1) flavor: be- Pasteurized whole milk (average protein-to-fat ratio ing overly ammoniacal, sulfury, or bitter, (2) texture: of 0.97; 25 kg/batch) was supplied by a regional milk having an overly runny paste, and (3) appearance: processor and transported to the Arbuthnot Dairy Cen- deterioration of rind color (browning). ter at Oregon State University (Corvallis) for cheese Manufacturers have modified cheese-making practices making. The milk was placed in a round cheese vat (C. to lengthen the window of optimum quality and create van’t Riet Dairy Technology B.V., Nieukoop, the Neth- stabilized bloomy rind cheeses (Lawrence et al., 1987). erlands) and warmed to the appropriate fermentation Stabilization of the paste is primarily accomplished by temperature for the recipe (Table 1). A thermophilic controlling the rate and level of acid development dur- starter culture, Streptococcus thermophilus (Choozit ing draining for a final pH after draining ≥5.2 (Gripon, DVI TA 50 series, Danisco, Copenhagen, Denmark), 1997). These modifications lead to a young cheese with or a mesophilic starter blend (Flora Danica-DVS, Chr. a softer and more elastic texture, but one that retains Hansen Inc., Milwaukee, WI), was added to the milk as physical integrity throughout ripening. Cheesemakers determined by recipe (Table 1). The ripening cultures have modified many manufacturing steps to produce [Penicillium candidum (PCA 3, Chr. Hansen Inc.), Ge- stabilized cheeses, including starter selection (i.e., lactic otrichum candidum (Choozit Geo 15 LYO, Danisco), acid bacteria species and strain choice), fermentation and Kluyveromyces marxianus (LAF 4, Chr. Hansen time/temperature, target rate and degree of acidifica- Inc.)] were added to the milk at a rate of 0.40, 0.15, tion, cut size, and curd handling practices (Lawrence et and 0.15 U/100 kg, respectively. When preparing the al., 1987; Gripon, 1997); however, data demonstrating stabilized curd variety, an adjunct culture, Leuconostoc the effect of these modifications on Camembert-type mesenteroides ssp. cremoris (Choozit LM 57, Danisco), cheese composition and quality characteristics through- was added at rate of 0.73 U/100 kg to enhance flavor out ripening are limited. development. Except for the lactic curd variety, cal-

Journal of Dairy Science Vol. 102 No. 1, 2019 166 BATTY ET AL. cium chloride (DCI Calcium Chloride, 32 to 33% wt/ vol, Dairy Connections Inc., Madison, WI) was added during fermentation at a concentration of 6.6 mL/100 kg. The milk was fermented until the desired set pH was achieved (Table 1). For all varieties, excluding the lactic curd, rennet None 6.6 9.7 5.3 3 stirs 6.2 6.45 1 5.2–5.3 40 40 (see concentration in Table 1; DCI Star Coagulant,

Stabilized curd Dairy Connections Inc.) was stirred gently into the fermented milk and the mixture set for 30 to 40 min before cutting with either 1- or 2-cm knives (Servi Do- ryl, Langeais, France). The cut time and size used for each cheese type is shown in Table 1. The cut time was determined by visual assessment of the gel firmness and flocculation time using a multiplication factor of 5. After cutting, the vats were drained and the forms (Fromagex, Rimouski, Québec, Canada; 7 cm in diam- None 6.6 3 + 8.8 5.3 3 stirs 6.0 6.2 1 4.9–5.0 eter) were filled using a curd distributor (Fromagex). 30 40 For the lactic curd variety, rennet was added once the Solubilized curd desired set pH was achieved (Table 1). Once the lactic

Mesophilic + thermophilic Thermophilic + adjunct curd variety achieved the desired cut pH, it was ladled into the cheese molds over the course of 2 to 3 h. Each mold was filled with curd every 30 to 45 min for 6 to 7 passes, until the molds were filled. Forty molds were filled for each replicate cheese make with an average cheese weight of 95 ± 10 g at salting. Cheeses were drained at 23 ± 1°C with turning at 6.6 9.7 5.3 3 stirs 6.0 6.2 2 4.8–4.9 30 35 30 min and 5, 10, and 19 h. The cheeses were removed Mesophilic from the molds at either 20 or 24 h and salted with whey with 35°C water All varieties were dry salted at a rate of 2% (wt/wt) were All varieties 2% salt (wt/wt) and open air-dried for an additional 1 to 2 h. This marked d 0 (start of ripening) for all downstream sampling points. Cheeses were transferred to an incubator (15 ± 1°C, 85% relative humidity) for 24 h. The relative humidity was then increased to None Replace 30% (vol/vol) 6.6 7.6 5.3 3 stirs 6.1 6.2 2 4.6–4.7 30 35 95% and temperature decreased to 13 ± 1°C for 10 d, and cheeses were flipped daily to develop surface mold. Cheeses were then wrapped in white mold paper (Fromagex) and stored at 7 ± 1°C for up to 40 d. After wrapping, cheeses were flipped on d 14, 21, 35, and 50 2

— postmanufacture. None None Ladle 7.6 1.6 4.6 6.0 4.2–4.4 22 Not added Mesophilic Mesophilic Lactic curd curd Sweet curd Washed Composition and Physiochemical Analysis Milk Composition. Compositional analysis (fat, protein, lactose, and SNF) of milk samples was con- ducted using a milk analyzer (Lacticheck-01, RapiRead, Hopkinton, MA) before culture addition on the day of cheese manufacture. Cheese Composition. Cheese samples were col- (mL/100 kg)

1 lected after 5 d of ripening (before development of sur- 3 face molds) for macrocomponent analysis (fat, protein, moisture, minerals). Whole cheese wheels (~90 g) were homogenized by blending until the size of the cheese quantity (mL/100 kg) quantity 2 particles were uniform (~15 s). Fat was measured ac- Rennet was added at a 1:40 dilution with deionized water. Rennet was letting the curd acidify and set for 24 h in vat. by achieved The lactic curd drain pH was cutting and hooping. the course of a 40-min holding time between of stirs over operations or curd handling: number Vat Starter quantity (U/100 kg) Starter quantity Drain pH Modifications in the cheese manufacture procedure to attain multiple variants of Camembert-type cheese of Camembert-type variants procedure to attain multiple manufacture 1. Modifications in the cheese Table Item CaCl 1 2 3 Starter type Fermentation temperature (°C) temperature Fermentation Set pH Cut size (cm) Washing pH at salting Rennet quantity Cut time (min) Vat operations Vat Salting cording to the Gerber Van Gulik method (ISO, 2008).

Journal of Dairy Science Vol. 102 No. 1, 2019 COMPOSITION AND CHARACTERISTICS OF CAMEMBERT-TYPE CHEESE 167 Protein was estimated using a combustion analyzer Color. The color of the rind was measured using a (vario MACRO cube, Elementar Analysensysteme spectrophotometer (LabScan XE Spectrophotometer, GmbH, Hanau, Germany) for total nitrogen determi- Hunter Associates Laboratory Inc., Reston, VA). All nation according to the Dumas combustion method color measurements were performed on 3 separate (Wiles et al., 1998). Crude protein was determined by cheeses for duplicate cheese makes of the rind through- multiplying total nitrogen by 6.38. Moisture content out ripening, 14, 21, 35, and 50 d postmanufacture. was determined using a rapid moisture and solids Paste Displacement. The stability of the cheese analyzer Computrac Max 4000XL (Arizona Instrument paste was measured quantitatively by measuring the LLC, Chandler, AZ). For elemental analysis (Ca, P, distance the paste traveled from the cut rind of the Na) 10 sub-samples were taken from each wheel and cheese wheel. On d 35 and 50, 3 cheeses for each cheese dried at 35°C for 96 h. The samples were then ground make were removed from 7°C storage, unwrapped, cut with a mortar and pestle. The samples were wet-ashed in half, and left undisturbed at ambient temperature with nitric acid (Macron Fine Chemicals, Center Val- for 60 min. Paste displacement (mm) was measured ley, PA) using microwave digestion (MultiwaveGO, An- from the edge of the cheese rind to the farthest point of tonPaar USA, Ashland, VA). Analysis was performed displacement after 15, 30, 45, and 60 min. by inductively coupled plasma atomic emission spec- Data Analysis. The data were analyzed using JMP troscopy (ICP-OES: Optima 2100 DV, Perkin Elmer, Pro 13.0 (SAS Institute Inc., Cary, NC). Means and Waltham, MA) operated using the radial view mode. standard errors were calculated for duplicate (n = 2) Total nitrogen and elemental analysis was conducted cheese makes for composition and shelf-life character- by the Collaborative Analytical Laboratory (Oregon istics. Comparison of means for significance (P < 0.05) State University, Corvallis). was conducted using Tukey’s honestly significant dif- Texture Analysis. Texture was analyzed by pen- ference. etrometery throughout ripening using a modified meth- od previously described by Abraham et al. (2007). A RESULTS texture analyzer TA XT2i (Texture Technologies Corp., Hamilton, MA) was equipped with a 5-kg load cell and Compositional Analysis a 6-mm cylindrical probe. The analysis was performed on 3 cheeses from 2 replicate cheese makes on d 7, 14, Milk. The average composition of the milk used to 21, 35, and 50 postmanufacture. The cheese was tem- manufacture Camembert-type cheese over 2 separate pered to room temperature (22 ± 1.5°C) for 1 h before days comprised of 3.57% fat, 3.48% protein, 5.05% lac- testing. Axially penetration was accomplished at a rate tose, and 9.22% SNF. The average protein-to-fat ratio of 0.4 mm/s with the penetration depth set at 75% of was 0.97. Protein and SNF of milk differed significantly the total cheese height. Firmness of the cheese paste (P < 0.05); however, protein-to-fat ratio of the milk did was recorded at the minimum peak and maximum peak not vary significantly. within the paste. The paste firmness was determined to Cheese Composition. Complete Camembert be the measurements observed after the initial fracture cheese composition for each variety is shown in Table peak that was due to the rind breaking. 2. Variation in the 5 recipes resulted in pH. The pH measurements for all in-process and significant differences in most of the compositional ripening samples were taken using a portable pH meter components analyzed in this study. The Na (dry basis; equipped with a conical penetration probe designed for 0.94–1.04%) was the only analyte that did not differ measuring semi-solid and solid food products (Portable significantly by recipe. Based on overall compositional Food and Dairy pH Meter, Hanna Instruments, Woon- comparisons, the 5 Camembert varieties can be orga- socket, RI). The pH of the rind was measured daily for nized in to 3 separate groups. These groupings are (1) the first 10 d of ripening and throughout total ripening lactic curd, (2) sweet curd and washed curd, and (3) (d 14, 21, 35, and 50). At each sampling time, 3 cheeses solubilized curd and stabilized curd. These groupings were measured in 6 locations. The sampling locations shared comparable manufacturing recipes. were arranged in a triangle pattern at an equal distance Compositional analysis demonstrated that the lactic apart on the top and bottom of the cheese wheel. The curd cheese was the most different from the other va- pH of the center paste was measured on d 7, 14, 21, rieties. The lactic curd cheese is acidified for up to 24 35, and 50 postmanufacture. The pH was measured for h, and is ladled in large pieces rather than cut, which 3 replicate cheeses for duplicate cheese makes through results in the highest moisture cheese (57.99%) with ripening. the lowest protein (15.43%), fat (21.75%), salt (1.05%),

Journal of Dairy Science Vol. 102 No. 1, 2019 168 BATTY ET AL.

Table 2. Composition expressed as % weight of cheese for 5 different varieties of Camembert-type cheese sampled 5 d after manufacture1

Component Lactic curd Sweet curd Washed curd Solubilized curd Stabilized curd Compositional grouping I II II III III Moisture 57.99 ± 0.60a 56.44 ± 0.74ab 56.29 ± 0.97b 53.28 ± 0.42c 53.15 ± 0.69c SNF2 20.59 ± 0.37b 19.90 ± 0.81b 20.79 ± 0.88b 23.96 ± 0.52a 23.50 ± 1.15a Fat3 21.75 ± 0.44b 23.66 ± 0.65a 22.92 ± 0.53ab 22.77 ± 0.79ab 23.35 ± 0.57a (50.98 ± 0.86bc) (54.33 ± 1.25a) (52.46 ± 1.27ab) (48.71 ± 1.37c) (49.88 ± 0.92bc) Protein3 15.43 ± 0.42c 15.90 ± 0.26c 16.62 ± 0.32b 18.23 ± 0.19a 18.34 ± 0.25a (36.72 ± 0.71b) (36.51 ± 0.66b) (37.16 ± 0.48b) (39.02 ± 0.39a) (39.15 ± 0.31a) Salt2 1.05 ± 0.06b 1.13 ± 0.11ab 1.12 ± 0.14ab 1.17 ± 0.12ab 1.24 ± 0.06a Salt-to-moisture (S/M)2 1.81 ± 0.10b 2.00 ± 0.19ab 2.01 ± 0.28ab 2.20 ± 0.21ab 2.33 ± 0.10a Na3 0.41 ± 0.02b 0.44 ± 0.04ab 0.44 ± 0.05ab 0.46 ± 0.04a 0.48 ± 0.02a (0.98 ± 0.06) (1.02 ± 0.09) (1.01 ± 0.11) (0.99 ± 0.11) (1.03 ± 0.04) Ca3 0.23 ± 0.02c 0.33 ± 0.04b 0.35 ± 0.03b 0.36 ± 0.02b 0.45 ± 0.03a (0.54 ± 0.06c) (0.75 ± 0.08b) (0.80 ± 0.05b) (0.77 ± 0.05b) (0.96 ± 0.06a) P3 0.21 ± 0.01c 0.32 ± 0.02b 0.31 ± 0.03ab 0.34 ± 0.01ab 0.40 ± 0.02a (0.51 ± 0.02b) (0.75 ± 0.05a) (0.72 ± 0.06a) (0.73 ± 0.03a) (0.86 ± 0.04a) a–cCheese composition values within a row that do not share the same superscript are significantly different (P < 0.05). 1Values presented are the means ± SE of 2 replicate cheesemakers (n = 2). 2Calculated values including SNF (100% − moisture % − fat %), Salt (Na × 2.54), S/M [(salt % × 100)/moisture %], % on an as-is basis, protein and minerals (Na, Ca, and P; {attribute × [(100 − moisture %)/100]}, % fat on a dry basis {[attribute/(100 − moisture %)] × 100}. 3Values in parentheses indicate component percentage on a dry basis. and salt-to-moisture (1.81%). The calcium and phos- Lactic Curd phorus contents of the lactic curd variety were the low- est (0.0.23 and 0.21%, respectively). Quality characteristics of lactic curd Camembert- No significant differences were observed in composi- type cheese during ripening are shown in Figure 1. The tion between the sweet and washed curd varieties ex- pH of the cheese at salting (d 0) was 4.31 (Figure 1a). cept for protein (15.90 and 16.62%, respectively). These Rind pH increased quickly achieving a pH of 6.98 by varieties are very similar in manufacture, the only ex- d 10 and a maximum pH of 7.92 on d 35. Increases ception being starter culture quantity and the washing in paste pH lagged, as expected, with an increase to of the curd. The washed curd was significantly lower 6.15 by d 21. By d 50, the pH of the rind and paste in moisture (56.29%) and it was significantly higher in harmonized at 7.65 to 7.70. The maximum firmness of protein (16.62%) compared with lactic curd. The sweet the lactic curd paste started low at just 1.13 N on d 7 curd was significantly higher in fat (23.66%) compared of ripening with little difference between minimum and with the lactic curd. The higher pH through the process maximum firmness (Figure 1b). Firmness continued to led to significantly higher (P < 0.05) calcium content in decrease until d 21 and held at a very low firmness these cheeses (sweet curd: 0.33%; washed curd: 0.35%) (0.34 N) through the end of the study (d 50). During compared with lactic curd with (0.23%). ripening there was a change in the color, primarily the The solubilized and stabilized curd cheeses had no lightness (L*) and yellow/blue coordinate (b*) values significant differences between the major compositional (Figure 1c). The L* value started at 87.19 on d 14, and components. Their manufacture differed from other va- by d 50 it had decreased to 68.58. The b* value saw an rieties by starter culture type, fermentation parameters, increase over from 6.58 (d 14) to a maximum of 14.56 and cut size which lead to cheeses with distinct char- (d 35). The red/green coordinate (a*) value changed acteristics. The moisture content of these 2 varieties only minimally from −0.43 to 1.42. Upon cutting the was significantly lower (53.15–53.28%) than the washed lactic curd wheel at the end of ripening (d 50), paste curd (56.29%) sweet curd (56.44%), and lactic curd quickly and almost completely left the rind of the lactic (57.99%) varieties. The protein content (wt/wt and curd cheese (Figure 1e). The paste of the lactic curd dry basis) of these 2 cheeses were the highest among variety spread quickly over a large distance, traveling the 5 varieties at 18.23–18.34% and 39.02–39.15%, re- 14 mm within 15 min after cutting (Figure 1d). The spectively. The calcium content of the solubilized curd average distance of paste displacement for the lactic (0.36%) was comparable to the sweet curd (0.33%) and curd variety was not significantly different at d 35 and washed curd (0.35%) varieties, whereas the calcium 50 d at 23.5 mm and 27.3 mm, respectively, after 60 content of the stabilized curd (0.45%) was significantly min of flow time. Taken together, these data indicate higher than all other varieties. A similar pattern was the rapid ripening of this cheese and demonstrate qual- seen for the phosphorous content. ity characteristics that explain its short shelf-life.

Journal of Dairy Science Vol. 102 No. 1, 2019 COMPOSITION AND CHARACTERISTICS OF CAMEMBERT-TYPE CHEESE 169

Figure 1. Characteristics of lactic curd Camembert variety from production throughout ripening (50 d). (a) pH of the rind and paste: rind (surface) pH (●), and paste (center) pH (○). (b) Firmness of the paste: maximum firmness (▲); minimum firmness (ripe zone, △). (c) Color of the surface represented by lightness (L*; ●), red/green coordinate (a*; ○), and yellow/blue coordinate (b*; △). (d) Paste displacement: 15, 30, 45, and 60 min. (e) Image of the paste displacement 50 d into ripening at the 60 min time point. Error bars represent the SEM for duplicate cheese makes (n = 2).

Journal of Dairy Science Vol. 102 No. 1, 2019 170 BATTY ET AL.

Sweet Curd Solubilized Curd Quality characteristics of sweet curd Camembert- Quality characteristics of solubilized curd Camem- type cheese during ripening are shown in Figure 2. This bert-type cheese during ripening are shown in Figure cheese, like the lactic curd, has a short ripening period 4. The pH at salting (d 0) started at 4.95 (Figure 4a), and shelf-life with substantial changes occurring be- with the rind pH reaching 7.29 by d 10 and a maximum tween 35 and 50 d as demonstrated by various quality of 7.48 by d 50. The paste pH increased to 5.04 by characteristics. The pH at salting (d 0) was 4.65 (Figure d 21 and 7.21 by d 50. Initial firmness (d 7) ranged 2a). The pH reached 7.38 by d 10 and the maximum of from 2.06 to 3.41 N (Figure 4b). The maximum and 7.85 by d 21. The pH of the paste increased slowly to minimum firmness continued to decrease until d 50, 5.27 by d 21 and reached a maximum of 7.53 by d 50. reaching 1.03 and 0.53 N, respectively. The L* value This cheese was quite firm with a maximum firmness of decreased over ripening like the other varieties, from 4.29 on d 7 of ripening (Figure 2b). Minimum firmness 87.57 at d 7 to 80.50 on d 50 (Figure 4c). The a* value of the paste was 1.91 on d 7. The difference between the exhibited minor fluctuation from −0.06 to 0.55. The b* maximum and minimum decreased throughout ripen- value also remained fairly constant ranging from 6.93 ing, harmonizing on d 35 at <0.80 N. The color of the to 7.56. Upon cutting the solubilized curd wheel at 50 sweet curd cheese rind slightly browned over time, with d, the paste remained closely associated with the wheel the L* value decreasing from 82.43 to 69.13 over ripen- even after 60 min from cutting (Figure 4e). The aver- ing (Figure 2c). The a* and b* values exhibited minor age distance of paste displacement for the solubilized changes, ranging from 0.24 to 1.00 and 6.35 to 8.78, curd variety was similar at d 35 and 50 at 4 and 6 respectively. Paste displacement significantly increased mm, respectively, after 60 min of flow time (Figure 4d). from d 35 to 50, averaging 11.7 mm on d 35 and 24.3 These quality characteristics indicate the potential for mm on d 50 after 60 min of flow time (Figure 2d and a longer shelf-life for the solubilized curd relative to the 2e). lactic, sweet, and washed curd variants.

Stabilized Curd Washed Curd Quality characteristics of stabilized curd Camembert- Quality characteristics of washed curd Camembert- type cheese during ripening are shown in Figure 5. The type cheese during ripening are shown in Figure 3. pH of the cheese at salting (d 0) was the highest of all The pH of the cheese at salting (d 0) was 4.80 (Figure varieties at 5.24 (Figure 5a). The rind pH increased to 3a). The rind pH increased rapidly to 7.37 by d 10, 7.10 by d 10 and reached a maximum by d 50 of 7.52. increased to a maximum of 7.81 by d 21, and remained The pH of the paste was slower to increase reaching above 7.62 through 50 d of ripening. The pH of the 5.30 by d 21 and 7.36 by d 50. The initial maximum paste was slower to increase, as expected, reaching 5.08 firmness of the stabilized curd was quite low at 2.51 N by d 21 and 7.60 by d 50. The maximum firmness of on d 7, whereas the minimum firmness was 1.70 on d 7 the washed curd paste was recorded on d 7 at 3.96 N, (Figure 5b). The firmness decreased over ripening, with with the minimum firmness on the same d at 2.14 N. the largest decrease happening between d 35 and 50. By d 35, the maximum firmness decreased to 0.65 N, The maximum firmness on d 50 was 1.35 N whereas the which was congruent with the minimum firmness of minimum was 0.78 N. Like the other varieties, the L* 0.45 N; low firmness remained constant throughout the value decreased over ripening from 83.75 to 77.58 (Fig- remaining ripening. Like the aforementioned described ure 5c). The a* value ranged from −0.16 to 0.76. The varieties, the L* decreased over ripening from 84.77 to b* value remained relatively unchanged, ranging from 74.07 (Figure 3c). The a* value ranged from 0.23 to 5.97 to 6.96. After the wheel was cut, the cheese paste 0.70 and the b* ranged from 6.72 to 8.22. Upon cut- moved very little over the 60 min as shown in Figure ting the washed curd wheel, the cheese paste moved 5e. The displacement distance after 60 min of flow time away from the cheese wheel at a lesser extent than the was 3 mm on d 35 and 4.3 mm on d 50 (Figure 5d). sweet and lactic curd, with much of the paste remain- Considering these physical quality characteristics, the ing in the wheel (Figure 3e). The displacement distance stabilized curd variety would have the longest shelf-life after 60 min of flow time significantly increased over of the varieties evaluated in this study. ripening, averaging 7.3 mm on d 35 and 18.7 mm on d 50 (Figure 3d). The difference in paste displacement DISCUSSION indicates that washed curd cheese undergoes much of the ripening between d 35 and 50, similar to the sweet Camembert recipe variations result in the production curd variety. of cheeses that differ substantially in their composi-

Journal of Dairy Science Vol. 102 No. 1, 2019 COMPOSITION AND CHARACTERISTICS OF CAMEMBERT-TYPE CHEESE 171

Figure 2. Characteristics of sweet curd Camembert variety from production throughout ripening (50 d). (a) pH of the rind and paste: rind (surface) pH (●), and paste (center) pH (○). (b) Firmness of the paste: maximum firmness (▲); minimum firmness (ripe zone, △). (c) Color of the surface represented by lightness (L*; ●), red/green coordinate (a*; ○), and yellow/blue coordinate (b*; △). (d) Paste displacement: 15, 30, 45, and 60 min. (e) Image of the paste displacement 50 d into ripening at the 60 min time point. Error bars represent the SEM for duplicate cheese makes (n = 2).

Journal of Dairy Science Vol. 102 No. 1, 2019 172 BATTY ET AL.

Figure 3. Characteristics of washed curd Camembert variety from production throughout ripening (50 d). (a) pH of the rind and paste: rind (surface) pH (●), and paste (center) pH (○). (b) Firmness of the paste: maximum firmness (▲); minimum firmness (ripe zone, △). (c) Color of the surface represented by lightness (L*; ●), red/green coordinate (a*; ○), and yellow/blue coordinate (b*; △). (d) Paste displacement: 15, 30, 45, and 60 min. (e) Image of the paste displacement 50 d into ripening at the 60 min time point. Error bars represent the SEM for duplicate cheese makes (n = 2).

Journal of Dairy Science Vol. 102 No. 1, 2019 COMPOSITION AND CHARACTERISTICS OF CAMEMBERT-TYPE CHEESE 173

Figure 4. Characteristics of solubilized curd Camembert variety from production throughout ripening (50 d). (a) pH of the rind and paste: rind (surface) pH (●), and paste (center) pH (○). (b) Firmness of the paste: maximum firmness (▲); minimum firmness (ripe zone, △). (c) Color of the surface represented by lightness (L*; ●), red/green coordinate (a*; ○), and yellow/blue coordinate (b*; △). (d) Paste displacement: 15, 30, 45, and 60 min. (e) Image of the paste displacement 50 d into ripening at the 60 min time point. Error bars represent the SEM for duplicate cheese makes (n = 2).

Journal of Dairy Science Vol. 102 No. 1, 2019 174 BATTY ET AL.

Figure 5. Characteristics of stabilized curd Camembert variety from production throughout ripening (50 d). (a) pH of the rind and paste: rind (surface) pH (●), and paste (center) pH (○). (b) Firmness of the paste: maximum firmness (▲); minimum firmness (ripe zone, △). (c) Color of the surface represented by lightness (L*; ●), red/green coordinate (a*; ○), and yellow/blue coordinate (b*; △). (d) Paste displacement: 15, 30, 45, and 60 min. (e) Image of the paste displacement 50 d into ripening at the 60 min time point. Error bars represent the SEM for duplicate cheese makes (n = 2).

Journal of Dairy Science Vol. 102 No. 1, 2019 COMPOSITION AND CHARACTERISTICS OF CAMEMBERT-TYPE CHEESE 175 tion and a variety of quality characteristics throughout The pH throughout the process controlled the final ripening. In this study, we produced Camembert-type total Ca in the cheese. Across the 5 varieties of Cam- cheeses using 5 different recipes. All recipes produced embert-type cheese the degree of acidification during cheeses that had a composition representative of Cam- the make varied significantly with the pH at salting embert-type cheese early in ripening (Schlesser et al., ranging from 4.3 to 5.3. Total Ca content trend for all 1992; Kulmyrzaev et al., 2005). The salting procedure cheeses was directly related to the pH at salting and or the alternative method used to calculate salt in this inversely related to the total degree of acidification. It study resulted in salt levels that were slightly lower has been previously reported that 3 main factors affect than expected; however, this was consistent across all demineralization during cheese manufacture including cheese varieties. Comparing the Na (dry weight basis) milk preacidification, pH during draining, and degree of content to that of the commercial cheeses studied by cooking (Lucey and Fox, 1993). In this study, multiple Tansman et al. (2017), the content was quite similar. methods were used to control pH and ultimately Ca The alternative processes (as compared with the tra- content. This included pH at rennet addition, quantity ditional lactic curd recipe) resulted in Camembert-type of starter culture added, type of starter culture used, cheeses with varying curd type, composition, and qual- fermentation temperature, and washing of the curd. ity. The traditional lactic curd cheese relies on acidifica- These process modifications limited the degree of acidi- tion as the catalyst for gel formation, whereas alter- fication during draining, and ultimately were successful native recipes limit acidification and drive enzymatic in controlling total retained Ca in the cheeses. Similar coagulation through the addition of rennet. Washed methods for controlling Ca content have previously curd, solubilized curd, and stabilized curd varieties had been used for mozzarella, Cheddar, and Colby cheeses an elevated pH compared with the other curd types, (Guinee et al., 2002; Upreti and Metzger, 2006; Lee which likely contributed to improved paste stability; et al., 2010). The cut size was also modified (1 cm however, only through the use of a thermophilic starter cubes) for the stabilized and solubilized as described (stabilized curd variety in this study) were we able to by Spinnler and Gripon (2004) when referring to the achieve the desired pH at salting for a stabilized type manufacturing methods for stabilized cheeses. This was cheese (of ≥5.2). Paste stabilization was also likely done to achieve an acceptable moisture content as the affected by the use of a mixed starter (thermophilic/ large cut size decreased whey expulsion during drain- mesophilic), and curd washing to limit acid production ing. This was likely due to the lesser degree of syneresis by the starter culture. occurring at a higher pH (Lund et al., 1971; Walstra, Shelf-life of Camembert cheese is largely a function of 1993). texture deterioration. Texture stability in Camembert This study was designed to characterize the com- cheese recipes was correlated with increasing calcium positional and quality characteristics of Camembert content in young cheeses. Lactic curd had the lowest cheeses produced by differing methods currently used initial total calcium content (0.23%) and resulted in the by Camembert-type cheesemakers. This study was least stable texture throughout shelf life. At 50 d, this not designed to systematically evaluate the effect of cheese paste nearly liquefied as demonstrated by the individual modifications; however, these results suggest lack of cheese remaining in the rind during the paste how modifications in specific cheese making processes displacement test (Figure 1e). Substantial improvement influence important cheese quality outcomes. It is also in texture stability was apparent when total Ca content important to note, that all cheeses in this study were of was increased, leading to Ca concentrations of ≥0.33%. a single diameter (7 cm). Camembert cheese diameter Calcium concentration in the young cheeses differed as and height will influence the rate of ripening and related a function of pH at salting and throughout the cheese quality characteristics; nevertheless, the mechanism for make. Higher pH at this stage and throughout the chee- bloomy rind and relative performance semaking process increases the insoluble colloidal cal- of recipes should be the same regardless of cheese size. cium phosphate content and overall interactions with casein (Keller et al., 1974; Lucey et al., 2003) leading CONCLUSIONS to a decrease in flowability and an overall more stable cheese paste. While all recipes with higher total Ca This study demonstrated significant differences in showed improved texture stability, the stabilized curd product composition as well as the pH changes, tex- with the highest pH at salting (pH 5.2–5.3) and highest ture/cheese body, and color development during ripen- total Ca content (0.45%) showed little softening and ing of 5 Camembert-type cheese variants. The study minimal paste displacement (flow) by d 50. The opti- provides the framework of different procedures that can mum time for consumption of this cheese would be >50 be used to make Camembert-type cheese and the effect d of ripening. these recipe variants will have on the quality and shelf

Journal of Dairy Science Vol. 102 No. 1, 2019 176 BATTY ET AL. life of the cheese. Understanding how different Cam- Karahadian, C., and R. C. Lindsay. 1987. Integrated roles of lactate, ammonia, and calcium in texture development of mold surface- embert-type cheese making recipes influence cheese ripened cheese. J. Dairy Sci. 70:909–918. https:​/​/​doi​.org/​10​.3168/​ composition will assist cheesemakers in controlling final jds​.S0022​-0302(87)80094​-2. product quality and allow them to manufacture cheese Keller, B., N. F. Olson, and T. Richardson. 1974. Mineral retention and rheological properties of Mozzarella cheese made by direct that will meet the desired expectations of the consumer acidification. J. Dairy Sci. 57:174–180. https:​/​/​doi​.org/​10​.3168/​ and requirements for their distribution network. jds​.S0022​-0302(74)84856​-3. Kulmyrzaev, A., É. Dufour, Y. Noël, M. Hanafi, R. Karoui, E. M. Qannari, and G. Mazerolles. 2005. Investigation at the molecular ACKNOWLEDGMENTS level of soft cheese quality and ripening by infrared and fluores- cence spectroscopies and chemometrics–Relationships with rheol- The authors thank the many parties that contrib- ogy properties. Int. Dairy J. 15:669–678. https:​/​/​doi​.org/​10​.1016/​ j​.idairyj​.2004​.08​.016. uted to the completion of this study. We thank Neville Lawrence, R. C., L. K. Creamer, and J. Gilles. 1987. Texture develop- McNaughton for the cheese recipe inspiration. We also ment during cheese ripening. J. Dairy Sci. 70:1748–1760. https:​/​/​ thank Yanyun Zhao’s laboratory members, especially doi​.org/​10​.3168/​jds​.S0022​-0302(87)80207​-2. Le Graet, Y., A. Lepienne, G. Brule, and P. Ducruet. 1983. Migration Zilong Deng and Jooyeoun Jung, for analytical support du calcium et des phosphates inorganiques dans les fromages à (Oregon State University, Corvallis). We very much ap- pâte molle de type Camembert au cours de l’affinage. Lait 63:317– preciate the help with cheese making and analysis from 332. https:​/​/​doi​.org/​10​.1051/​lait:​1983629​-63019. Leclercq-Perlat, M.-N. 2011. Camembert, Brie, and Related Variet- Eva Kuhn, Danica Berry, Alex Emch, Robin Frojen, ies. Encycl. Dairy Sci. 773–782. https:​/​/​doi​.org/​10​.1016/​B978​-0​ and the Arbuthnot Dairy Center (Oregon State Univer- -12​-374407​-4​.00091​-1. sity, Corvallis) team. Funding for this study was sup- Lee, M.-R., M. E. Johnson, S. Govindasamy-Lucey, J. J. Jaeggi, and J. A. Lucey. 2010. Insoluble calcium content and rheological proper- ported by Dairy Management Inc. (Rosemont, IL) and ties of Colby cheese during ripening. J. Dairy Sci. 93:1844–1853. the Charles E. and Clara Marie Eckelman Fellowship. https:​/​/​doi​.org/​10​.3168/​jds​.2009​-2734. Lucey, J. A., and P. F. Fox. 1993. Importance of calcium and phos- phate in cheese manufacture: A review. J. Dairy Sci. 76:1714–1724. REFERENCES https:​/​/​doi​.org/​10​.3168/​jds​.S0022​-0302(93)77504​-9. Lucey, J. A., M. E. Johnson, and D. S. Horne. 2003. Invited Review: Abraham, S., R. Cachon, B. Colas, G. Feron, and J. De Coninck. Perspectives on the basis of the rheology and texture properties 2007. Eh and pH gradients in Camembert cheese during ripening: of cheese. J. Dairy Sci. 86:2725–2743. https:​/​/​doi​.org/​10​.3168/​jds​ Measurements using microelectrodes and correlations with tex- .S0022​-0302(03)73869​-7. ture. Int. Dairy J. 17:954–960. https:​/​/​doi​.org/​10​.1016/​j​.idairyj​ Lund, D. B., N. F. Olson, and M. C. Patel. 1971. Factors affecting .2006​.12​.010. syneresis of renneted milk gels. J. Dairy Sci. 55:913–918. https:​/​/​ American Cheese Society. 2016. State of the US Artisan/Spe- doi​.org/​10​.3168/​jds​.S0022​-0302(72)85596​-6. cialty Cheese Industry. Accessed Jan. 14, 2018. http:​/​/​ Schlesser, J. E., S. J. Schmidt, and R. Speckman. 1992. Characteriza- 1sskha48hhbhl0oc61bhsagh​.wpengine​.netdna​-cdn​.com/​wp​ tion of chemical and physical changes in Camembert cheese during -content/​uploads/​2017/​02/​2016​-State​-of​-the​-US​-Artisan​-and​ ripening. J. Dairy Sci. 75:1753–1760. https:​/​/​doi​.org/​10​.3168/​jds​ -Specialty​-Cheese​-Industry​-Summary​-Report​.pdf. .S0022​-0302(92)77934​-X. American Cheese Society. 2018. ACS & Industry Growth. Accessed Shaw, M. B. 1981. The manufacture of soft, surface mould, ripened Jan. 14, 2018. http:​/​/​www.cheeseso​ ciety.org/​ ​membership/​the​-acs​ cheese in France with particular reference to Camembert. J. Soc. -universe/.​ Dairy Technol. 34:131–138. Bouma, A., C. A. Durham, and L. Meunier-Goddik. 2014. Start-up Sousa, M. J. 2003. Cheeses|Surface Mold-ripened Cheese Varieties. and operating costs for artisan cheese companies. J. Dairy Sci. Encycl. Food Sci. Nutr. 1122–1129. https:​/​/​doi​.org/​10​.1016/​B0​-12​ 97:3964–3972. https:​/​/​doi​.org/​10​.3168/​jds​.2013​-7705. -227055​-X/​00807​-5. Buragas, A. 2006. Brie, Camembert Bring Royal Heritage to U.S. Cheese Spinnler, H.-E., and J.-C. Gripon. 2004. Surface mould-ripened chees- Market. Accessed Dec. 19, 2017. http:​/​/​www.cheesemark​ etnews​ es. Cheese: Chemistry. Physics and Microbiology. 2:157–174. .com/​articlearch/​cheese/​27oct06​_02​.html. Tansman, G. F., P. S. Kindstedt, and J. M. Hughes. 2017. Crystal- Galli, B. D., J. G. P. Martin, P. P. M. da Silva, E. Porto, and M. H. lization and demineralization phenomena in stabilized white mold F. Spoto. 2016. Sensory quality of Camembert-type cheese: Rela- cheese. J. Dairy Sci. 100:6074–6083. https:​/​/​doi​.org/​10​.3168/​jds​ tionship between starter cultures and ripening molds. Int. J. Food .2016​-12259. Microbiol. 234:71–75. https:​/​/​doi​.org/​10​.1016/​j​.ijfoodmicro​.2016​ Upreti, P., and L. E. Metzger. 2006. Influence of calcium and phospho- .06​.025. rus, lactose, and salt-to-moisture ratio on Cheddar cheese quality: Gripon, J. C. 1997. Flavour and Texture in Soft Cheese. 2nd ed. B. Manufacture and composition. J. Dairy Sci. 89:420–428. https:​/​/​ A. Law, ed. Springer Science & Business Media, Berlin, Germany. doi​.org/​10​.3168/​jds​.S0022​-0302(06)72106​-3. Guinee, T. P., E. P. Feeney, M. A. Auty, and P. F. Fox. 2002. Effect Walstra, P. 1993. The syneresis of curd. Cheese Chem. Phys. Micro- of pH and calcium concentration on some textural and functional biol. 141–191. https:​/​/​doi​.org/​10​.1007/​978​-1​-4615​-2650​-6​_5. properties of mozzarella cheese. J. Dairy Sci. 85:1655–1669. https:​ Wiles, P. G., I. K. Gray, and R. C. Kissling. 1998. Routine analysis of /​/​doi​.org/​10​.3168/​jds​.S0022​-0302(02)74238​-0. proteins by Kjeldahl and Dumas methods: Review and interlabo- International Organization for Standardization (ISO). 2008. Cheese. ratory study using dairy products. J. AOAC Int. 81:620–632. Determination of fat content. Van Gulik method. ISO Standard Wisconsin Milk Marketing Board. 2017. 2017 Dairy Data. Accessed 3433. International Organization for Standardization, Geneva, Jan. 9, 2018. http:​/​/​www​.wisconsindairy​.org/​assets/​images/​pdf/​ Switzerland. WisconsinDairyData​.pdf.

Journal of Dairy Science Vol. 102 No. 1, 2019